The 18th International Conference on Behçet's Disease De Doelen, Rotterdam, Netherlands, 13-15 September 2018

The 18th International Conference on Behçet’s Disease
                De Doelen, Rotterdam, Netherlands, 13–15 September 2018

The 18th International Conference on Behçet’s Disease took place in the Dutch port city of
Rotterdam. The programme included three keynote lectures, 22 presentations from ISBD invited
speakers, 10 oral presentations of accepted abstracts, 140 e‐posters, one debate and three
sponsored symposiums.

Mihai Netea (Netherlands) started proceedings with a keynote lecture entitled ‘Innate immunity’.
The innate immune system provides the first line of defence from infection in a non‐specific manner.
Unlike adaptive immunity, which responds more slowly to specific threats, the innate system does
not build memory. Although improved response on a second exposure (the basis of vaccination) is an
evolutionary advantage, only vertebrates (about 5% of all species) have an adaptive immune system,
whereas innate immune systems are found in all classes of plants and animals. However, some plants
and invertebrates that survive infections become more resistant. Mice with a deficiency in adaptive
immunity can develop some protection against future infections. This is known as trained immunity,
a form of innate immune memory with features distinct from classical adaptive immune memory. It
involves reprogramming of myeloid cells, with activation of the cholesterol synthesis pathway. A
metabolite called melavonate is important in this process. Monocytes of patients with hyper
immunoglobulin D syndrome (HIDS), who are deficient in mevalonate kinase and accumulate
mevalonate, have a constitutive trained immunity phenotype. There is also evidence that BCG
vaccination against tuberculosis provides non‐specific protection against other infections through a
training effect on monocytes and bone marrow progenitor cells. Finally, trained monocytes produce
pro‐inflammatory cytokines and may contribute to chronic inflammation.

Basic science

As usual, the first session of the conference was devoted to basic science. Andre Uitterlinden
(Netherlands) spoke about the complex genetics of common diseases. Human DNA is highly variable,
with more than 150 million variable loci contributing to the necessary heterogeneity of the

population. The Rotterdam Study has been running for 25 years and includes around 20,000 people
aged over 45. With more than 2 billion data points, this is an example of ‘big data’ and is examining
how DNA changes as a result of environmental factors. Osteoporotic fractures, for example, have a
complex phenotype. Risk factors include bone strength, fall risk and impact force, and these are
influenced by genetic factors (DNA) and dynamic genomics (RNA, epigenetics, DNA methylation, etc).
The populations of Holland and Belgium have similar genetics but a different diet, with different
calcium intake affecting bone strength, for example. Monogenic diseases, such as cystic fibrosis and
sickle cell disease, arise from mutation in a single gene and can be investigated by whole genome
sequencing. The common diseases have more complex genetics and are investigated by genome‐
wide association studies (GWAS), which look for associations with genetic variants. A study of
140,000 people in the UK Biobank found genetic variants that accounted for about 20% of the
heritability of bone mineral density. The cost of DNA analysis has decreased in recent years, and the
time has come to start using it as a clinical tool. It is also necessary for researchers to collaborate and
publish meta‐analyses of large amounts of data rather than multiple publications of small amounts of
data.

Eun Ha Kang (Korea) then gave a presentation on the genetic aspects of Behçet’s disease (BD) in Asia.
The strongest evidence is for HLA‐B51, with the prevalence of BD in a population largely depending
on the prevalence of HLA‐B51. This association is lower in Korea than in the rest of Asia. Another
association, especially in B51‐negative people, is with HLA‐A26. This association is higher in Japan
than in Korea, and is not seen in Chinese people. Two MICA polymorphisms are also important – one
of these seems to be associated with HLA‐B51 while the other is independent. MICA has a role in the
pathogenesis of BD. TNFα also plays a central role in BD, but the TNFα promoter polymorphism has a
lower association in Asia than in other populations. In addition, a GWAS in Korea did not find any
association with the IL10 and IL23R‐IL12RB2 variants seen elsewhere.

Next, Ivona Aksentijevitch (USA) spoke about BD‐like diseases and their genetic association with BD.
There are several rare monogenic diseases with similarities to BD. A family in Canada has several
members diagnosed with BD, but the inheritance is dominant and the family members are HLA‐B51
negative. Two candidate genes have been identified – tumour necrosis factor (TNF) receptor and
TNFα‐induced protein‐3 (TNFAP3). Another family (also HLA‐B51 negative) has systemic lupus
erythematosus (SLE) and BD‐like features, and a mutation in the TNFAP3 gene has been identified.
Low‐prevalence single‐nucleotide polymorphisms (SNPs) in this gene are associated with multiple
autoimmune diseases, including SLE, psoriasis, type 1 diabetes, inflammatory bowel disease (IBD)
and rheumatoid arthritis (RA), as well as with protection against asthma and allergies. A20 is an anti‐
inflammatory protein that regulates cell death and has a tumour suppressive function. Patients with
haploinsufficiency of A20 (i.e. only one copy of the gene is functional) have several BD‐like features.
Onset is usually in childhood, and fever, autoantibodies and polyarthritis are more common than in
BD.

To round off the session, Ahmet Gül (Turkey) gave an overview of the genetics of BD and the
consequences for disease phenotypes. HLA‐B51 and ERAP1 are the most important genetic factors in
defining the phenotype, while other factors are important as triggers of disease. The distribution of
HLA‐B51 in populations is similar to the distribution of BD. B51 has a strong association with ‘classic’
symptoms of BD such as mucocutaneous, joint and eye symptoms, but less association with
gastrointestinal, vascular and neurological manifestations. HLA‐B15, B27 and B57 are also risk alleles,

while HLA‐A26 confers risk in Asian populations. There are also protective HLA alleles, including B49
and A03. B51‐positive people can also have polymorphisms in ERAP1. ERAP1 haplotype 10 is
particularly associated with BD, while haplotypes 1 and 3 are associated with ankylosing spondylitis
(AS) and psoriasis. The link between ERAP1 haplotypes and hyperinflammatory responses has not yet
been elucidated. However, it is clear that genetic susceptibility combined with environmental
triggers lead to a hyperinflammatory response involving the innate and adaptive immune systems,
which results in the clinical manifestations of BD and similar diseases.

In the first oral presentation of an abstract, Elaine Remmers (USA) reported on a whole‐genome
imputation study looking for new genetic loci contributing to BD.1 The researchers used data from
the original Turkish GWAS study, including 311,459 markers in 1278 BD patients to impute genotypes
of over 5.9 million markers. They identified a marker in the TLR2 locus that seems to confer risk for
BD by increasing the innate immune response against microbes. The morning was completed by
Maria Padula (Italy), who presented results from a study of ERAP1 polymorphisms in BD and AS in
Italian patients.2 Using DNA from 51 BD patients and 53 AS patients, the group found a different
distribution of the most common ERAP1 coding variants in the two diseases. In particular, the
rs17482078 genotype was more common in the BD patients; this SNP is thought to be associated
with risk of BD but protective against AS.

The basic science session continued in the afternoon, with a presentation on microbiomes from
Marcel de Zoete (Netherlands), who described the microbiota as the forgotten organ. Each
microbiota is unique and is composed of 100 to 200 bacterial species, as well as 150 times more
genes than human cells. The microbiota affects health and disease in many ways, being involved in
obesity, colon cancer, IBD, metabolic syndrome and possibly autism. Patients with IBD have been
shown to have an altered (dysbiotic) microbiota, with some species being absent and some extra
species being present. The hypothesis is that an ‘inflammatory’ member of the microbiota causes low
level inflammation that leads to chronic disease. Immunoglobulin A (IgA) secreted into the gut
neutralises bacteria and can be used to identify the species by staining and sequencing the IgA‐
coated bacteria. This approach has been validated in mice and has been used to show a lot of extra
IgA‐coated bacteria in the faeces of IBD patients. Healthy volunteers have fewer IgA‐coated bacteria,
and a different pattern. Cultures of IgA‐positive bacteria from IBD patients trigger severe colitis in
germ‐free mice, while IgA‐negative cultures trigger mild colitis; IgA‐positive cultures from healthy
people trigger only mild colitis. IgA‐positive Bacillus fragilis has been shown to adhere to intestinal
cells and to be enriched in enzymes that degrade mucus. BD patients have more IgA‐positive
Bifidobacteria species than healthy people, and this could represent a future target for treatment.

Giacomo Emmi (Italy) then spoke about the role of the microbiome in BD. He explained that short‐
chain fatty acids such as butyrate have an important role in maintaining gut health. BD patients have
decreased diversity of gut microbiota species, with depletion of genera that produce butyrate. The
salivary microbiota is also less diverse; immunosuppressant treatment increases some species, but
improving oral health does not increase diversity. However, in Japan, no difference in diversity was
seen between BD patients and healthy controls, and immunosuppressants had no effect on diversity.
Vegetarians have more butyrate‐producing species and decreased inflammatory markers.
Approaches to improving the microbiota include prebiotics (compounds that induce the growth or
activity of beneficial bacteria) and probiotics (live bacteria), as well as postbiotics which contain
bacterial metabolites such as butyrate. A lacto‐ovo vegetarian diet and supplementation with

butyrate both increase the ability of plasmin to degrade thrombus, which decreases the risk of
inflammation‐induced thrombosis.

Next, Haner Direskeneli (Turkey) gave a presentation on epigenetics, which involves changes in
phenotype without genetic changes, brought about by environmental factors such as ageing or
smoking. One example is DNA methylation, which suppresses gene expression; it is decreased in BD
and can be restored by colchicine treatment. Hypermethylation of the promoter region of the
interleukin (IL)‐10 gene leads to reduced production of IL‐10 (an anti‐inflammatory cytokine) in BD
patients, while hypomethylation of the IL‐6 promoter region results in increased IL‐6 (a pro‐
inflammatory cytokine). Ocular BD is associated with aberrant methylation of IRF8, a transcription
factor that plays a critical role in myeloid cell maturation. Another epigenetic mechanism involves
micro RNAs, which are small, non‐coding RNA molecules involved in RNA silencing and post‐
transcriptional regulation of gene expression. There are more than 900 miRNAs in mammals, and
they are affected by inflammatory responses. For example, miR155 is decreased in the peripheral
blood mononuclear cells and dendritic cells of BD patients with active uveitis. Some miRNAs are
increased in BD, while others are decreased, and miRNA profiling in BD patients shows many affected
genes and pathways, such as neutrophils, cell adhesion and oxidative stress. However, most of these
studies have not been replicated, and the functional effects have not been studied. DNA methylation
inhibitors have been investigated in some cancers, but have not yet been looked at in
autoinflammatory diseases.

Two more oral presentations of abstracts followed. First, Seonghyang Sohn (Korea) showed data on
the chemokine receptor CCR1 and its ligand CCL3 in a mouse model of BD,3 suggesting the
upregulation of CCR1‐positive cells is related to the control of systemic inflammation in BD. However,
an audience member questioned the validity of mouse models in BD. Harry Petrushkin (UK) then
presented interesting research on the role of the KIR gene cluster, which co‐evolved with HLA and is
expressed in innate immune cells.4 This work used DNA from 267 UK BD patients and 445 healthy
controls. The combination of low expressing KIR3DL1 allotype and KIR3DS1 was associated with
ocular BD, whereas high KIR3DL1 and null KIR3DR1 provided protection against mucocutaneous
disease. This research may offer insight into the pathogenic role of HLA‐B51 and its interaction with
KIR3DL/S1.

The first day of the conference finished with a special lecture by Graham Wallace (UK) on the
biomechanisms of BD. He began by pointing out that the association of HLA‐B51 with BD is well
known, but the exact role of B51 is not well understood and many BD patients are B51‐negative. B51
is a very old protein, so it must be important to have been conserved during evolution. In addition to
B51, and other common polymorphisms in IL‐10 and IL‐12R/23R, more than 45 SNPs have been
found to be associated with BD, some of them only in particular parts of the world. For example,
PTPN22 (protein tyrosine phosphatase, non‐receptor type 22) is found in 18% of BD patients in
northwestern Europe but is not seen in the Far East. However, many of the different genes identified
may be involved at different places in the same pathways. For example, PTPN22 in Europe and
GIMAP in the Far East are both involved in lymphocyte regulation. This means that it is important to
look at pathways rather than individual genes. A similar situation is seen with the influence of the
microbiome and dietary on BD. Certain foods trigger oral ulcers in BD patients, with different foods
triggering the same pathways (e.g. histamine release) in different countries. Metabolomic analysis is
the large‐scale study of small molecules (metabolites), within cells, biofluids, tissues or organisms. A

clear separation is seen between patients with BD and early arthritis, and this may help to identify
pathways for further investigation. Research is trying to link polymorphisms to metabolomics
differences in different parts of the world. Dr Wallace concluded by saying that greater
understanding is needed of the role of HLA‐B51, the cells involved in BD, and the role of the
microbiome and the diet.

Nature versus nurture

The second day began with a debate entitled “Nature or nurture: does migration change the
prevalence of BD?”. Yohei Kirino (USA) began by making the case for the role of nature, laying out six
reasons why BD can be considered a hereditary disease. These are prevalence depending on
ethnicity, the association with HLA‐B51, identification of identical IL‐10 and IL‐23R SNPs in Turkey and
Japan, familial cases of BD, reduction of heterogeneity by population stratification (e.g. ERAP1), and
the evolving genotype in Japan (and other phenotype clusters, such as ocular and neurological
manifestations versus gastrointestinal and vascular symptoms). Haner Direskeneli (Turkey) then
presented a case for the role of the environment, pointing out that migration changes the prevalence
of BD. The prevalence of BD is decreasing worldwide, with good data from Korea showing a
reduction from 7.4 to 2.5 per 100,000 between 2006 and 2015. In Japan, the rate of HLA‐B51
positivity in BD patients is also decreasing, and BD is evolving to become more like IBD. It is likely that
changes in phenotype are related more to environmental factors than to genetics. The discussion
that followed unsurprisingly concluded that both nature and nurture are important. Environmental
factors alone will not trigger BD in a person with no genetic tendency. Shigeaki Ohno (Japan), who
has looked at BD among descendants of Japanese emigrants in Hawaii and Brazil, pointed out that
the incidence of BD is very low in these groups, although their prevalence of HLA‐B51 is the same as
in Japan. It may be that the early immigrants had a high prevalence of BD, but after three or
generations the effect of the environment has reduced the incidence. It is notable that the incidence
of other genetic diseases is the same as in Japan. It was agreed that environmental factors are
important in determining the severity of BD, but that a genetic predisposition is essential for the
disease to develop.

The oral presentations of abstracts continued with Yutaro Soejima presenting the data on clustering
of BD types in Japanese patients that had been discussed in the preceding debate.5 Analysis of 691
patients identified three subgroups: group A includes ocular and neurological manifestations, group
B includes gastrointestinal and vascular symptoms, and group C patients have primarily
mucocutaneous symptoms (oral/genital ulcers and skin lesions). The group A cluster has declined in
incidence since the 1990s, while the B cluster has increased. It was pointed out that gastrointestinal
symptoms are a minor feature in the Middle East and are not included in the international criteria for
BD. Gülem Hatemi (Turkey) had presented the main results of a clinical trial of apremilast in BD (207
patients) at the 2016 conference, showing a significant effect on oral ulcers. This time, she showed
the results for patient‐reported outcomes, including measures of disease activity and quality of life.6
Significant improvements were seen in all the measures, showing that the reduction in oral ulcers
with apremilast was associated with decreased disease activity and increased quality of life. To finish
the session, Farhad Shahram (Iran) reported on 204 paediatric cases from the Iran registry of BD.7
These cases, 2.7% of the 7504 patients in the registry, represent the largest paediatric BD cohort

reported. The mean age of disease onset was 10.5 years. Ocular lesions were more common than
reported in other cohorts, while genital ulcers, skin lesions and gastrointestinal involvement were
less common.

Organ involvement in BD

The first clinical session began with a talk on large vessel vasculitis by David Saaduin (France), who
explained that BD can involve both arteries and veins of all sizes. Vascular manifestations include
deep vein thrombosis (DVT), pulmonary artery aneurysm (PAA)/thrombosis and dural sinus
thrombosis. Vascular infiltrates include neutrophils, T cells and NT cells, and imaging shows
inflammation of the vessel wall. There is geographical variation in the incidence of vasculo‐BD, and
both incidence and severity are higher in male patients. Large vessel vasculitis tends to be recurrent.
The most common vascular manifestation is venous thrombosis, especially DVT in the legs. Aortic
aneurysm is the most common arterial manifestation. Superficial thrombophlebitis, caused by a
blood clot just below the surface of the skin, is a risk factor for DVT and must be recognised and
treated. Cerebral venous thrombosis represents 30% of neuro‐BD and 10% of vasculo‐BD; it is
strongly associated with DVT in the legs. Budd–Chiari syndrome is a rare but severe presentation that
is often associated with inferior vena cava thrombosis; early treatment improves the prognosis.
Another manifestation that must be diagnosed as soon as possible is pulmonary embolism, which is
caused by pulmonary venous thrombosis and associated with pulmonary artery thrombosis and
aneurysm. The new EULAR recommendations say that vasculo‐BD should be treated with
corticosteroids and immunosuppressive drugs, with anti‐TNF agents in refractory cases and possible
use of anticoagulants. In DVT, immunosuppression and anticoagulation have been shown to be
better than anticoagulation alone, while PAA is best treated by embolization and
immunosuppression. Anti‐TNF therapy can produce regression of PAA, avoiding surgery and reducing
the risk of relapse. Overall, early diagnosis and aggressive treatment can improve the prognosis of
vasculo‐BD.

Farida Fortune (UK) then spoke about oral manifestations, emphasising their effect on quality of life,
ability to work and psychosocial functioning. Examination of the mouth is not well taught in medical
schools, and many dentists are not trained in clinical medicine. The mouth is complex, including
several different types of tissue in the tongue, hard palate and cheeks. On top of all the tissue is the
saliva, which is different from other mucosa in the body (e.g. gut or vagina), while the salivary glands
are different from other glands. Periodontal disease is important, with Gram‐negative bacteria being
involved, while different organisms are found in the saliva of BD patients with active ulcers. Oral
ulcers occur in 15–25% of the population and in 30–40% of relatives of BD patients. They can be
caused by many factors, such as trauma, infection, nutritional deficiency or cancer, and patients can
be wrongly diagnosed with BD. One distinction is that the outside of the lips is not affected in BD.
The ulcers in patients with Crohn’s disease are similar to those in BD, whereas ulcers seen in
ulcerative colitis patients are quite different. SLE patients also have distinctive ulcers with ‘star
flaming’. The 2014 criteria allow patients to be diagnosed with BD on the basis of having both oral
and genital ulcers, but they may not have BD. It is important that the unique features of BD are
recognised.

The session continued with a presentation on the eye by Ilknur Tugal‐Tutkun (Turkey). He said that
uveitis is a result of intraocular inflammation, and 90% of cases of uveitis in BD are in males. BD
uveitis can be both anterior and posterior; it is usually bilateral, although isolated anterior uveitis
occurs (about 10% of cases, often female). Around 90% of BD uveitis has posterior segment
involvement, and a minority of patients have conjunctival ulcers and other manifestations. BD uveitis
typically manifests as recurrent acute episodes, which can be explosive; spontaneous resolution is
common. Macular ischaemia can lead to permanent visual loss, while chronically leaky retinal
capillaries increase the risk of macular oedema. The differential diagnosis of BD uveitis includes
several other diseases, and BD patients can also develop uveitis from other causes. Fundus
fluorescein angiography is useful in the differential diagnosis, while optical coherence tomography is
a quick and non‐invasive method of visualising the characteristic features of retinal infiltrates in BD.
Colour fundus photography shows the transient nature of retinal infiltrates. Work is ongoing to
define and validate criteria for the diagnosis of ocular BD.

Next, Gulsen Akman‐Demir (Turkey) spoke about neurological aspects of BD, pointing out that not
every neurological symptom in BD patients represents neuro‐BD. A thorough history and
examination, including magnetic resonance imaging (MRI) and analysis of cerebrospinal fluid (CSF),
are needed to make a diagnosis. About 70% of neuro‐BD is parenchymal, involving the brainstem
and/or the spinal cord; another 20% is represented by dural sinus thrombosis (presence of a blood
clot in the dural venous sinuses, which drain blood from the brain), and 10% have atypical multiple
sclerosis (MS)‐like symptoms. The CSF is usually abnormal in parenchymal disease and normal in
dural sinus thrombosis, and the prognosis is better in the latter. Particular care needs to be taken
regarding complications of treatment and of common co‐morbidities in patients with atypical neuro‐
BD. Care is also needed in patients with typical neuro‐BD symptoms who do not fulfil the criteria for
BD. Headache in BD patients is usually migraine or tension headache rather than a symptom of
neuro‐BD. Neuro‐BD can present with headache, however, but other neurological findings are
needed for the diagnosis to be made. MS‐like neuro‐BD should be treated more like MS than BD, and
anti‐TNF agents should be avoided. Any MRI features that do not respond to steroids are not neuro‐
BD. There is no class I evidence for treatment of neuro‐BD. The recommended approach is high‐dose
intravenous methylprednisolone for an acute attack, followed by tapering of the steroid and
introduction of immunosuppressive therapy. First‐line immunosuppressants are the conventional
drugs such as azathioprine, with interferon or a TNF inhibitor in refractory patients.

The last presentation in this session was on the skin and was given by Bing Thio (Netherlands).
Mucosal lesions (skin lesions and oral or genital ulcers) are not life‐threatening and do not cause
serious damage, but they reduce quality of life. They are also important in the diagnostic criteria for
BD; 97% of patients have oral ulcers, and 75% have genital ulcers. The microbiome plays a large role
in these ulcers; it has been proposed that a ‘disease‐associated microbial community’, deficient in
butyrate‐producing bacteria, can induce epithelial barrier damage, allowing entry of damaging
molecules. Autoinflammatory processes in the skin involve neutrophils, macrophages and IL‐1.
Manifestations such as pyoderma gangrenosum and erythema nodosum are typical of BD. The risk of
post‐thrombotic syndrome (redness, swelling, ulcers and chronic leg pain following DVT) may also be
increased. Treatment of skin lesions in BD usually involves topical corticosteroids, colchicine and
azathioprine, with biologics such as TNF inhibitors and the IL‐1 inhibitor apremilast also having a role.

‘Capita selecta in BD’

This session included a variety of clinical topics, starting with Eldad Ben‐Chetric (Israel) speaking
about pregnancy and contraception in BD patients. Most people with BD are in their reproductive
years, so this is an important issue. Both the effect of pregnancy and contraception on BD and vice
versa are matters of concern. Some retrospective case series have been published, but very few
include comparisons with controls. In one series, four out of six cases reported a decrease in
mucocutaneous symptoms with oral contraceptive use, while the other two worsened. In case
reports of 32 pregnancies in BD patients, 14 had reduced disease activity, 16 worsened and two were
unchanged; it cannot be excluded that these findings were simply due to chance. In 15 case series
including 1164 pregnancies, 37% improved, 23% worsened and 40% were unchanged. However,
there is a possibility that cases of deterioration are more likely to be reported. In addition, some
patients have a different disease course in different pregnancies. If disease exacerbation occurs, it
can be in any trimester. In 11 case series, outcomes of pregnancy were similar to those in healthy
women, with a possible tendency to increased rates of miscarriage and caesarean section. The risk of
miscarriage may be related to thrombotic events, while caesarean sections may be due to fetal
distress or risk of perineal trauma.

Annet van Rooyen (Netherlands) then spoke about BD in children, which accounts for between 3%
and 36% of cases. The average age of onset is 7–8 years, and oral ulcers are the first symptom in
80%. BD in children is more likely to be familial than it is in adults. It is usually 2–3 years before the
second symptom appears, which can delay diagnosis. However, there is a large overlap with other
autoinflammatory diseases, so it is important not to rush to a diagnosis. Among a cohort of 219
children with possible BD in 12 countries, 156 have a confirmed diagnosis. Fever is reported as a
symptom in almost 44%, which is very different from BD in adults. Genital ulcers are less common
than in adults (55% versus 80%), as are ocular symptoms. This cohort has been used to develop new
PEDBD criteria, in which fulfilment of 3/6 items represents BD. All suspected BD in children should be
documented and followed up to confirm the diagnosis. Treatment of paediatric BD is similar to that
in adults, with short courses of steroids, immunosuppressants such as azathioprine and
mycophenolate mofetil, TNF inhibitors and IL‐1 or IL‐6 inhibitors. Clinical trials never include children,
but this needs to be reconsidered so that data can be collected.

Next was a presentation on cardiovascular risk factors and surgery in BD patients, given by ISBD
President Dorian Haskard (UK). Any operative procedures in BD patients carry a risk of a pathergy
reaction, including insertion of cannulas and catheters. Dental procedures can trigger oral ulcers,
while general and cardiovascular surgery may lead to postoperative aneurysms or saphenous vein
graft thrombosis. Vascular surgery can cause graft thrombosis, aneurysm, vessel occlusion or fistula
formation. Gastrointestinal surgery has a 50% recurrence rate and can cause perforation, fistula
formation and obstruction. Poor surgical outcomes in BD result from amplified inflammatory
responses. There is ‘cross‐talk’ between inflammation and thrombosis, which are linked by tissue
factor. Tissue factor‐bearing microparticles in blood plasma increase the thrombogenic potential,
with a positive feedback loop leading to increased inflammation. In addition, corticosteroids,
immunosuppressant drugs and biologics increase the risk of infection and impair wound healing. It is
nevertheless important to suppress inflammation as much as possible before planned surgery, but
this is more difficult for emergency surgery.

The second day of the conference was completed by the final three oral presentations of abstracts.
Tom Missotten (Netherlands) described a study comparing interferon and anti‐TNF therapy in a
cohort of 52 BD patients with uveitis.8 Of the 100 eyes, 31 were treated mainly with anti‐TNF agents
(mostly adalimumab) and 12 were treated mainly with interferon. The 10‐year visual outcomes did
not differ between the two groups, and neither did the inflammatory activity. Fatma Alibaz‐Oner
(Turkey) then presented data on venous vessel wall thickness in the legs in male BD patients.9
Venous thrombosis due to vascular inflammation is the most common form of vascular involvement
in BD, and this study showed that vascular wall thickness was increased in 61 male BD patients (30
with vascular involvement) compared with 37 healthy controls and 27 patients with AS. If it can
predict future vascular involvement, this may be a useful diagnostic tool. The final presentation of
the day was by Amal Senusi (UK), on the association of alpha‐melanocyte stimulating hormone (α‐
MSH) and vasoactive intestinal peptide (VIP) with fatigue and quality of sleep in BD.10 The study
included 100 BD patients and 30 healthy controls, and found that both α‐MSH and VIP were higher in
the BD patients. α‐MSH (as well as IL‐6) was associated with both fatigue and poor sleep quality,
while VIP was associated with poor sleep quality and disease activity. A better understanding of
these complex interactions might lead to the development of novel approaches to manage fatigue
and sleep disorders as well as disease activity in BD patients.

New therapeutic strategies

The final morning of the conference was dedicated to treatment of BD. The first presentation, by
Marjan Versnel (Netherlands), was on targeting interferon type I and II pathways in autoimmune
diseases. Type I interferons include interferon alpha and beta, while interferon gamma is type II.
Type I interferons are produced by plasma B cells and activated T cells, monocytes and macrophages.
A ‘signature’ of interferon type I‐induced gene expression has been identified in Sjögren’s syndrome,
allowing patients to be classified as type I‐positive or negative. Type I‐positive patients have
increased disease activity compared with negative patients. The signature is not specific to Sjögren’s
syndrome, and type I‐positive SLE and RA patients have also been identified. A type I receptor
blocker is being tested in SLE; 34% of patients met the primary endpoint in a phase 2 study, and a
phase 3 trial is ongoing. A type II signature has also been identified in Sjögren’s syndrome, and some
patients are positive for both types. Type I interferonopathies are rare monogenic
immunodeficiencies, and the TBK1 (TANK binding kinase‐1) pathway (involving TLR7/9 signalling) has
been identified as playing an important role. TBK1 is increased in type I‐positive patients with
Sjögren’s syndrome, SLE and systemic sclerosis; a TBK1 inhibitor is a potential treatment. Janus
kinase (JAK) inhibitors block both type I and II interferons and are a potential treatment for patients
positive for both signatures. The JAK/STAT pathway is activated in BD, and type I interferon is used as
a treatment for BD. Research is needed to look at interferon signatures in BD and determine whether
subgroups based on these might be useful in selecting and monitoring treatment.

Martin van Hagen (Netherlands) then spoke about biosimilars in BD, explaining that while generic
forms of small molecule drugs are identical to the original drug, biosimilars are highly similar, but not
identical, to the original biologics. Biologics are defined as proteins containing more than 40 amino
acids and manufactured using DNA technology. They are a mix of structurally related isoforms, with
differing glycosylation, and even the reference products change over time. In some cases, a

biosimilar may look more like the original reference product than the current reference product
does. An infliximab biosimilar was approved for use in RA, spondyloarthritis and IBD in 2016. No
clinical trials of biosimilars have been done in rare diseases. One observational study looked at 48
patients with rare autoinflammatory diseases (12 with BD) who were switched from infliximab
originator (Remicade) to biosimilar (Remsima). Two patients switched back due to adverse events;
after 6 months, another six had switched back and five had stopped treatment. No differences in
various efficacy and safety measures were seen (including BD activity), but two patients with
sarcoidosis experienced seizures and two had worsening polyneuropathy. A systematic review of 57
studies of biosimilars in any disease identified important evidence gaps around the safety of
switching between biologics and their biosimilars. A report of three BD cases documented rapid
disease relapse after switching from Remicade to Remsima, possibly due to cross‐reaction of anti‐
infliximab antibodies.

Next, Peter Sfikakis (Greece) gave an update on anti‐TNF therapy, concentrating on timing of
tapering. TNF inhibitors, especially infliximab, are the key biologic drugs used in BD and have been
used since 2001 to treat BD uveitis. Tapering of therapy is possible, as some early patients have
remained relapse‐free for 10 years when treatment was stopped after 3 years. Infliximab has been
shown to be effective in long‐term treatment of uveitis, with 41% of patients not relapsing during 4
years’ treatment, and it has also been shown to be more effective than adalimumab. Earlier use of
infliximab achieves better outcomes, and it is a useful therapy for gastrointestinal, vascular and
neurological manifestations of BD as well as uveitis. A Japanese study found that safe, pre‐planned
discontinuation of infliximab therapy can be performed in patients with Behçet’s uveitis. In a study of
infliximab therapy in severe ocular BD, 11/13 patients responded and achieved long‐term remission
for a mean of 7 years; four of them were drug free. Dr Sfikakis has also published a study showing
that drug‐free, long‐term remission after withdrawal of successful anti‐TNF treatment is feasible in
patients with severe BD. Of 46 patients on long‐term anti‐TNF treatment, 29 discontinued and 12 of
them remained in remission for at least 3 years. Of the other 17 patients, 14 were re‐treated and
four these subsequently achieved drug‐free remission. It seems that younger patients with shorter
disease durations are more likely to achieve long‐term, drug‐free remission.

Finally, Gülem Hatemi (Turkey) summarised the use of molecular designed agents in BD, beginning
with TNF inhibitors. These are well established in the treatment of BD, especially refractory ocular
symptoms. Good effects have also been reported in patients with vascular problems such as
pulmonary artery involvement. A recent study evaluating anti‐TNF therapy in refractory BD patients
with major vessel involvement reported that vascular remission was achieved in 16/18 patients.
Among patients with gastrointestinal BD symptoms who received anti‐TNF therapy in another study,
clinical remission was obtained in 47/91 (51%) patients, while endoscopic remission was observed in
21/46 (45%). TNF inhibitors have also been shown to be effective in the treatment of neuro‐BD. Most
of these studies used infliximab, and sometimes adalimumab, but recent studies have also reported
good results with golimumab and certolizumab. Immunogenicity (production of antibodies against
the anti‐TNF drug) seems to be less of a problem in BD than in RA or Crohn’s disease. Another class
of biologic drugs that has been used in BD is the IL‐1 inhibitors anakinra, canakinumab and
gevokizumab. These have been shown to be effective against oral/genital ulceration; they also have
some effect in uveitis, but much less than is seen with TNF inhibitors. The IL‐17 inhibitor
sekukinumab did not reduce disease activity in BD patients. In contrast, the IL‐12/23 inhibitor
ustekinumab brought about complete remission in 9/14 BD patients with oral ulcers refractory to

colchicine. Finally, tocilizumab, an IL‐17 inhibitor, has also been shown to be effective in refractory
BD, particularly for neurological involvement and uveitis.

The ethics of genetic analysis

To conclude the conference, James Stacy Taylor (a moral philosopher) gave a keynote lecture on the
ethics of genetic analysis, stressing that he could only provide a framework for discussion and not
absolute answers. Questions arising from genetic testing include whether patients have a right not to
know about incidental findings, and whether they have a responsibility to disclose information to
other people affected by it. The principles at play here are respect for autonomy, beneficence, non‐
maleficence and justice. In a case where only the patient is affected, there is no difficulty, but such
cases are rare. In most cases, patients may have a moral obligation to disclose information to others
affected, which may mean that they do not have the right not to know. A duty of disclosure may
deter some people from being tested. It may also depend on whether the disease is treatable; many
people would prefer not to know if they have a degenerative disease that cannot be treated. A useful
framework (which can be used by ethics committees) is to identify the issue, determine whose
interests deserve moral consideration, and decide which of the above principles can be used to
support the decision.

Closing remarks

ISBD President Dorian Haskard closed the conference, noting the important themes of big data and
improvements in the treatment of BD with biologics. The hope for the future is that these two
aspects will come together to provide individualised treatment, ensuring the right treatment for the
right patient at the right time. It was announced that the 19th ICBD will take place in Athens on 2–4
July 2020.

                                                                                                    Clare Griffith

References
1. Remmers F, Takeuchi M, Meguro A, et al. Whole‐genome imputation study implicates TLR2 locus variants
that confer risk for Behçet’s disease by increasing innate immune response against microbes. (O01) Clin Exp
Rheumatol 2018; 36 (Suppl. 115): S15.
2. Padula MC, Leccese P, Lascaro N, et al. A comparative study about ERAP1 polymorphisms in Behçet
syndrome and Ankylosing spondylitis Italian Patients. (O02) Clin Exp Rheumatol 2018; 36 (Suppl. 115): S153.
3. Sayeed HM, Sohn S. Chemokine Receptor CCR1 and its ligand CCL3 in a mouse model of Behcet’s disease.
(O03) Clin Exp Rheumatol 2018; 36 (Suppl. 115): S154.
4. Petrushkin H, Norman P, Wallace G, Stanford M, Parham P, Fortune F. The combination of a low expressing
KIR3DL1 allotype and KIR3DS1 contributes toward the risk of developing Behçet’s Disease. (O04) Clin Exp
Rheumatol 2018; 36 (Suppl. 115): S154.
5. Soejima Y, Kirino Y, Takeno M, et al: Clustering analysis of Japanese Behçet’s disease identifies intestinal type
as distinct cluster. (O05) Clin Exp Rheumatol 2018; 36 (Suppl. 115): S154.

6. Hatemi G, Mahr A, Ishigatsubo Y, et al. Disease activity and quality‐of‐life improvements in patients with
Behçet’s syndrome: a phase III randomized, placebocontrolled, double‐blind study of apremilast (RELIEF). (O06)
Clin Exp Rheumatol 2018; 36 (Suppl. 115): S155.
7. Shahram F, Nadji A, Akhlaghi M, et al. Pediatric Behcet’s Disease; Report of 204 cases from the Iran registry
of Behcet’s disease. (O07) Clin Exp Rheumatol 2018; 36 (Suppl. 115): S155.
8. Missotten T, van Velthoven MEJ, Hermans ME, et al. Behcet’s Uveitis : Comparison of Interferon‐a vs. anti‐
TNF‐a therapy in a clinical cohort. (O08) Clin Exp Rheumatol 2018; 36 (Suppl. 115): S156.
9. Alibaz‐Oner F, Ergelen R, Mutis A, et al. Venous vessel wall thickness in lower extremity is increased in male
Behcet’s Disease patients. (O09) Clin Exp Rheumatol 2018; 36 (Suppl. 115): S156.
10. Senusi A, Liu J, Bevec D, et al. The Association of Alpha‐Melanocyte Stimulating Hormone, Vasoactive
Intestinal Peptide with Fatigue and Quality of Sleep in Behçet’s Disease. (O10) Clin Exp Rheumatol 2018; 36
(Suppl. 115): S196.